Christian Kaltepoth

IMO your understanding is correct and both Jersey and RESTEasy doesn't behave correctly. But it would be interesting to hear other thoughts about this as well.

I basically like the idea. IMO creating something like a `UriTemplate` class representing a template and providing methods to match a concrete URI against the template to extract the named...

Partially related: The Chrome team [announced in early November](https://groups.google.com/a/chromium.org/g/blink-dev/c/K3rYLvmQUBY/m/vOWBKZGoAQAJ?pli=1), that they will remove Server Push from their HTTP/2 and HTTP/3 protocol implementations. See also this [blog post](https://evertpot.com/http-2-push-is-dead/) for some context....

I think that this is a sign that Server Push isn't widely adopted, and therefore I don't think that working on an API for this feature makes a lot of...

I also like the idea of having improved support for common data types like `LocalDate`. Of course the spec already enforces that implementations support a static `valueOf()` method. However, I...

> The spec could mandate that each compliant implementation MUST provide an implementation of `ParamConverterProvider` with lowest `@Priority` (to be able to override it) that provides an instance of `ParamConverter`...

Wouldn't it make sense to allow the user to access the annotations via the `ResourceInfo` interface? Something like: ```java public interface ResourceInfo { // other methods Annotation[] getResourceMethodAnnotations(); } ```...

I'm just guessing here, but as SSE doesn't define a way to include the content type in an event, the media type was most likely just added to the JAX-RS...

Just to fully understand this. The proposal is to not only explicitly allow to split implementation jars into client and server but also to define how APIs must behave if...

@spericas Ok. I got this impression from the following formulation: > If the client capabilities are not included but are used, the following APIs will fail in the **following manner**....